Why does sampling error occur?

Answer 1

The list is very long! In sampling error, I include all aspects of data collection. Samples (and not the full population) are taken in observational and experimental studies. The sample size may be a problem. In some cases, it may impossible to correct. If I am studying some rare occurrence, say hurricanes with winds over 280 mph or incidences of mad cow disease, the number of observations is fixed. Other times, there is inadequate time or budget to sample adequately. Sampling error can occur because of the way a sample is taken. This is very true of marketing surveys, which may be taken at time when they are more likely to survey one segment of the population. Or they are taken in one location that is not representative of the general population. A voluntary survey, or convenience survey may also be biased. The manner in which questions are posed, can introduce bias. Inadequate quality checking also contributes to sampling error. This is true whether the data collection is done by humans, or instruments, such as a testing laboratory. If a particular instrument is improperly calibrated, all measurements can be questionable. Finally, there are many means of purposely introducing bias into collected data in order to show "factual evidence" of preconceived ideas. The time frame or location where data is to be collected may be done to build in a particular bias. You will probably find more examples by searching the internet.